🏆
Rankings
LLM Leaderboard
Real measurements of every tested language model on real, named hardware – rated by raw speed (performance in tokens per second, prefill and time to first token) and by practical task quality in complete agent and chat runs (harness). Pick a benchmark type below or filter by model, maker and hardware to see exactly what is tested and how the results are produced.
⚡ Performance (tok/s)🤖 Harness quality👥 Concurrency🖥️ real hardware
| # | Model / Maker | Metrics | Parallel | GPU / CPU / RAM | Runtime | ||
|---|---|---|---|---|---|---|---|
| 1 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 148,79 tok/s TG Prefill 3.362 · TTFT 6.588 ms | 10× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 2 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 123,54 tok/s TG Prefill 2.897 · TTFT 3.743 ms | 5× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 3 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 101,36 tok/s TG Prefill 3.237 · TTFT 6.850 ms | 10× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ8_0 | Details → | |
| 4 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 84,61 tok/s TG Prefill 2.849 · TTFT 3.806 ms | 5× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ8_0 | Details → | |
| 5 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 76,95 tok/s TG Prefill 2.356 · TTFT 929 ms | 1× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ4_K_M | Details → | |
| 6 | Nemotron-Cascade-2-30B-A3B30BNVIDIA Performance benchmarkTimebench 3 - Kombi (Prefill + Generation) | 62,04 tok/s TG Prefill 2.341 · TTFT 935 ms | 1× | NVIDIA GB10 (DGX Spark)NVIDIA Grace · 120 GB RAM | llama.cppopenclaw_cliQ8_0 | Details → |